Poll Position

This is the first publicly released poll I’ve seen on the Himes v. Debicella race. It comes from an outfit called the American Action Forum. Its findings?

Despite challenger Dan Debicella having just 35 percent name recognition compared to Rep. Jim Himes’s 93 percent name recognition, Himes leads by a slim 46 to 42 percent margin.

That’s weird.

The AAF polled a number of congressional races around the country and found the GOP candidates doing well. The methodology, as reported, does not seem too peculiar:

All respondents were selected randomly from a list of registered voters in the district, and indicated they are likely to vote in the elections for Congress this fall, and interviews were conducted by live interviewers. Quotas in each district… were set by gender, age, and county consistent with past participation in the district. Each district included 400 interviews.

But without cross-tabs, it’s difficult to see just how representative this sample is. I notice party affiliation is not one of the quotas listed. I’m no expert on polling, but I would assume that’s the single biggest variable in poll results, so it seem to me that a truly representative sample ought to reflect party enrollment district-wide. Then again, if they truly sampled 400 randomly off the voter registration list, I think the sample should be pretty proportional to the general population. That would be true of gender or age as well, so why stipulate that those are representative by design but omit party affiliation? Hmmm.

I am skeptical because I don’t believe that only 12% of voters in our district are undecided. I also don’t believe 42% are going for Debicella yet. Plus, the fivethirtyeight.com pollster ratings don’t even list AAF, so I figured they must be brand-new. Indeed, they are. Brand new and GOP-branded. From a NY Times article announcing the formation of AAF last February, in an article titled “Republican Group to Promote Conservative Ideas”:

A group of prominent Republicans is forming an organization to develop and market conservative ideas, copying a successful Democratic model and hoping to capitalize on the fund-raising and electioneering possibilities opened up by a recent Supreme Court ruling.

Republicans who are donors, board members or both include Haley Barbour, the governor of Mississippi; Jeb Bush, former governor of Florida; Ed Gillespie, like Mr. Barbour a former chairman of the Republican Party; Fred Malek, an investor and official in the Nixon and first Bush administrations; Robert K. Steel, a former executive of Wachovia and Goldman Sachs who was a Treasury official in the second Bush administration, and Kenneth G. Langone, a founder of Home Depot and a former director of the New York Stock Exchange.

Maybe their polling is not suspect. Maybe they aren’t trying to advance a narrative or manipulate coverage of the race. And maybe people in CT-04 know the names of all the state senators in the district (quick — who represents Oxford in state senate? Yeah, I don’t know either) and feel like they have enough information to cast a ballot. Or maybe this poll is garbage, being promoted by a right-wing outfit through right-wing media to give Debicella a fundraising talking point.

I can’t really tell.

Update: A reader sent me the cross-tabs! Turns out they are available online here if you care to flip through 104 pages of questions and responses. If I’m reading it correctly, the poll included 22% Republicans, 30% Democrats, 44.5% unaffiliated, and a handful of non-responders or minor party respondents. Yet, by “ideology” 39.5% identified as Conservative, 24% Liberal, and 33.5% Moderate (the remainder must not have responded or chose something else).

And, notably, 44% of those polled expressed support for the Tea Party. I don’t believe 44% of likely voters in our district support the Tea Party. No way.